End of Course Tests: High Stakes Exit Exams
States, including Georgia, are moving toward greater use of End of Course Tests
(EOCTs) as high school exit exams. According
to a report published on August 13th, only four states used some form
of EOCT as a graduation requirement in 2007. By 2015, at least 11 states will
use EOCTs as a high school exit requirement. While the list of 11 states does
not currently include Georgia, state law (O.C.G.A.
§ 20-2-281) calls for the Georgia High School Graduation Test to be
replaced by EOCTs "according to a schedule to be determined by the state
board," and the state is already moving in this direction.
EOCTs are tied more closely to classroom instruction than are other types of
exit exams, and currently in Georgia, EOCTs count for 15 percent of a student's
final course grade. With the importance of EOCTs set to grow, it may be
worthwhile to examine the relationship between grading practices and EOCT
results. Both measures are based on the same state curriculum. A lack of
alignment between the two - if a significantly greater percentage of students
passed a course than passed the EOCT, for example - could indicate grade
inflation. Research suggests a growth in grade inflation nationwide since the
early 1990s. Grade point averages seem to be rising faster than other forms of
student achievement,
such as the ACT. In Arizona,
a newspaper investigation found that many students who could not pass
state assessments were awarded passing grades in their classes.
Given 1) the growing importance of EOCTs; 2) that EOCTs, because they
immediately follow classroom instruction, should show outcomes that are
somewhat aligned to course grades; and 3) that Georgia has embarked upon a
complete overhaul of its curriculum, it is important to ask: How do Georgia's
school systems fare when their high school grading alignment practices are
examined?
Grading Alignment in Georgia High Schools
Recently, the Governor's Office of Student Achievement (GOSA) commissioned Dr.
Christopher Clark, an Economics professor at Georgia College and State
University, to study the alignment of local school systems' grading practices
with their EOCT results. Using data obtained from the 2006-07 school year, Dr.
Clark analyzed the alignment between Georgia students' EOCT performance levels
and the grades they earned in related courses. Why is the relationship between
the two important? His findings suggest that these disparities may impact: a)
college success, b) HOPE Scholarship retention rates, and c) students' need for
learning support (i.e. remedial classes) in college.
A Look at the State-Level Grading Alignment
Each of the following three tables represents a different aspect of Dr. Clark's
results for the state of Georgia as a whole. Similar tables containing results
for each school system in which more than 100 students were tested are posted
under the research tab on GOSA's website:
http://www.gaosa.org/research.aspx. (Detail on the study's methodology
can be found in the study itself on GOSA's website.)
Table 1: Statewide EOCT Exceeds Rates by Course Grade
The table below shows the percentage of students exceeding standards (i.e.
scoring at the highest level) on each EOCT by the grade they received in each
course. For example, 76.98% of the students statewide who earned an A in
Algebra I also exceeded standards on the Algebra I EOCT. This table also shows
that 8.38% of students who failed their Economics course actually exceeded
standards on the Economics EOCT.
Table 2: Statewide EOCT Meets + Exceeds Rates by Course Grade
The second
table shows the percentage of students who either met or exceeded standards on
each EOCT by course grade. For example, 94.03% of the students statewide who
earned an A in Algebra I also met or exceeded standards on the Algebra I EOCT:
Table 3: Statewide EOCT Failure Rates by Course Grade
The last table
shows the same data in Table 2 from the opposite perspective: the percent of
students who failed each EOCT, by course grade. For example, 5.97% of the
students statewide who earned an A in Algebra I also failed the Algebra I EOCT.
In this case, one might reasonably expect that students who earned high course
grades had low failure rates on the EOCTs.
What Does the Data Look Like for Individual School Systems in Georgia?
To arrive at a numerical value to represent each school system's grading
practices across subject areas, GOSA compared students who earned high course
grades AND high EOCT scores against students who only earned high course
grades. GOSA divided the total number of students who earned an A, B, or C in
each course AND who exceeded standards on the EOCT by the total number of
students who earned an A or B in the course. (i.e. [A+B+C students who exceeded
EOCT standards] ÷[A+B students]). It would be reasonable to assert that such
school systems whose students exceeded standards on the EOCTs and who earned an
A, B, or C in the course have high grading practice integrity. This ratio is
presented below as the "Grade Alignment Score."
Using the method, if GOSA scored systems based only on the EOCTs in which
they had 100 or more test takers on at least one EOCT, the list of top 20 looks
like the table below (158 systems):
Using the same method, the table below shows the 20 systems with at least 100
test takers on all eight EOCTs (95 systems) who earned the highest score
for their course grading practices.
 Looking at this data in other
ways provides slightly different pictures of school systems' grading practices.
For example, some school systems consistently had a high percentage of students
who earned C course grades but who exceeded standards on EOCTs. In other words,
even the "average" students in these systems consistently scored well on
statewide standardized assessments in multiple subject areas.
The table to the right shows school systems that were consistently among the
top 20 systems statewide in terms of the percentage of their C students who
exceeded standards on each EOCT. This table lists the systems and the number of
EOCTs (out of 8) on which they had a top-20 ranking in the percentage of C
students scoring at the Exceeds performance level. These systems have grading
practices such that even students with average grades score well on
standardized tests.
Conversely, a few systems consistently assigned "A" grades to students who
scored at the Did Not Meet (DNM) standards performance level on the EOCTs. This
would indicate inflated grades across subject areas. The systems who landed in
the top 20 statewide in terms of the percentage of A students who scored DNM on
5 or more EOCTs were Atlanta Public Schools, Clayton County Public Schools, and
Crawford County Schools. As stated above, EOCTs count for 15 percent of a
student's course grade. This means that poor performance on an EOCT leads
directly to a lower grade; therefore, the part of the grade determined by
teachers may be even more inflated than these data suggest.
Conclusion
The conclusion to Dr. Clark's research paper best sums up the information
provided above. He found that "there are considerable grading disparities
across Georgia's High School Algebra, English Literature, Biology, Physical
Science, History, Geometry, and Economics classes. Comparing student's course
grades to their End of Course Test (EOCT) scores indicates that some school
systems appear to be inflating course grades relative to the EOCT scores
considerably while others appear to hold their students to higher standards.
These disparities are disconcerting because they may impact college success,
HOPE scholarship retention rates for HOPE scholars, and the need for learning
support (remedial classes) in college. Students from schools and school systems
that appear to consistently inflate grades may be less likely to succeed in
college courses, less likely to retain the HOPE scholarship, and more likely to
need to take remedial classes after enrolling in college than students from
schools and school systems that hold their students to higher standards.
Future research should be undertaken to analyze the impact of grading
disparities on later academic success. An examination of the impact that rigor
in grading standards (or a lack of rigor) may have on a student's academic
future should be performed once data on HOPE eligibility, HOPE retention rates,
and performance in college courses become available for the students whose data
were used in this study."
For the full research report, please visit
http://www.gaosa.org/research.aspx.
|
